Post date: Jun 10, 2017 3:41:9 AM
These should get you everything:
/uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_wings/Gemma/output_all/pip*txt
/uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_wings/Gemma/output_all/eff*txt
/uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_wings/Gemma/output_grps/pip*txt
/uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_wings/Gemma/output_grps/eff*txt
The pips* files give the posterior inclusion probabilities and the effe* files give the model averaged effect estimates. They both have the same general format. There is one row per SNP. The first two columns provide the scaffold and SNP position. The remaining columns give the PIPs or effect estimates for each trait (one trait per column). So, if you wanted to plot pips for the first areas in L. anna (as an example), the R code would be:
pips<-read.table("pips_AN_size.txt",header=FALSE)
plot(pips[,3]) ## 3 not 1 because the first two columns are the scaffold and position
And if you wanted to plot the first 6 areas on a single plot with different colors for each it would be something like:
pips<-read.table("pips_AN_size.txt",header=FALSE)
cs<-c("orange","blue","gray","brown","violet","forestgreen")
plot(pips[,3],col=cs[1],ylim=c(0,1))
for(i in 2:6){
j<-i+2 ## 2 offset accounts for first two columns
points(pips[,j],col=cs[i])
}
And you could plot the first 16 size pips on a single page with 16 panels like this:
pips<-read.table("pips_AN_size.txt",header=FALSE)
par(mfrow=c(4,4)) ## sets up a 4 x 4 panel plot = 16 panels
for(i in 1:16){
j<-i+2 ## 2 offset accounts for first two columns
plot(pips[,j])
}
The number of SNPs is not necessarily the same for different subgroups or for size and position. This is because SNPs that are fixed or nearly fixed are automatically dropped and this could apply to different SNPs for different subgroups. We can cross-reference them if you want, and we probably will want to at some point, but just plotting the individual ones first might be useful.